28
Santa Clara Law Santa Clara Law Digital Commons Faculty Publications Faculty Scholarship 4-29-2011 Domestic Application of Treaties David Sloss Santa Clara University, [email protected] Follow this and additional works at: hp://digitalcommons.law.scu.edu/facpubs Part of the Law Commons is Book Chapter is brought to you for free and open access by the Faculty Scholarship at Santa Clara Law Digital Commons. It has been accepted for inclusion in Faculty Publications by an authorized administrator of Santa Clara Law Digital Commons. For more information, please contact [email protected]. Automated Citation David Sloss, Domestic Application of Treaties (2011), Available at: hp://digitalcommons.law.scu.edu/facpubs/635

Domestic Application of Treaties

  • Upload
    others

  • View
    1

  • Download
    0

Embed Size (px)

Citation preview

Santa Clara LawSanta Clara Law Digital Commons

Faculty Publications Faculty Scholarship

4-29-2011

Domestic Application of TreatiesDavid SlossSanta Clara University, [email protected]

Follow this and additional works at: http://digitalcommons.law.scu.edu/facpubsPart of the Law Commons

This Book Chapter is brought to you for free and open access by the Faculty Scholarship at Santa Clara Law Digital Commons. It has been accepted forinclusion in Faculty Publications by an authorized administrator of Santa Clara Law Digital Commons. For more information, please [email protected].

Automated CitationDavid Sloss, Domestic Application of Treaties (2011),Available at: http://digitalcommons.law.scu.edu/facpubs/635

Electronic copy available at: http://ssrn.com/abstract=1826102Electronic copy available at: http://ssrn.com/abstract=1826102

Domestic Application of Treaties

David Sloss

INTRODUCTION

There has been dramatic growth in treaty making since World War II: more than 44,000

treaties were registered with the United Nations between 1945 and 2007.1 Meanwhile, with the

rise of globalization, the boundary separating domestic from international law has become

increasingly permeable. Consequently, states are making greater use of treaties to regulate

activity that was previously regulated exclusively by domestic law. For example, under the 1993

Hague Convention on Intercountry Adoption,2 eighty-three states have agreed to regulate child

adoption on a transnational scale.3 Additionally, states are concluding greater numbers of treaties

that protect the rights of private parties, including, for example, treaties related to international

human rights law,4 international humanitarian law,

5 and international refugee law.

6 As a

consequence of these three trends — growth in the number of treaties, increasing overlap

between treaties and domestic law, and a growing emphasis on private rights — domestic courts

are playing an increasingly prominent role in treaty application.

Traditional scholarship on the domestic application of treaties has focused on the

distinction between monist and dualist legal systems.7 Part One of this chapter explains that

distinction: in brief, the monist-dualist divide hinges on the role of the legislative branch in

incorporating and implementing treaties domestically. Although the monist-dualist framework

helps illuminate important formal differences among states, Part One suggests that scholarly

preoccupation with the formal distinction between monism and dualism tends to obscure key

functional differences among states.

Hence, the remainder of the chapter adopts a functional approach, focusing primarily on

the role of domestic courts in promoting compliance with treaty obligations and protecting

treaty-based private rights. Part Two explains the distinction between horizontal, vertical and

transnational treaty provisions. Part Three addresses the functional distinction between

nationalist and transnationalist approaches to judicial application of treaties. Part Four discusses

the crucial role of domestic courts in promoting compliance with treaty obligations, especially

transnational and vertical treaty obligations.

Professor of Law, Director of the Center for Global Law and Policy, Santa Clara University School of Law.

1 See United Nations Treaty Series Cumulative Index <http://treaties.un.org/Pages/CumulativeIndexes.aspx>

accessed 29 December 2010. In contrast, states concluded about 16,000 treaties during the nineteenth century. See

John Fabian Witt, ‗Internationalism and the Dilemmas of Strategic Patriotism‘ (2006) 41 Tulsa L. Rev. 787, 791. 2 Convention on Protection of Children and Cooperation in Respect of Intercountry Adoption (adopted 29 May

1993, entered into force 1 May 1995) 1870 UNTS 167. 3 See Status Table, Convention on Protection of Children and Cooperation in Respect of Intercountry Adoption

<http://www.hcch.net/index_en.php?act=conventions.status&cid=69> accessed 29 December 2010. 4 See, e.g., International Covenant on Civil and Political Rights (adopted 16 December 1966) 999 UNTS 171.

5 See, e.g., Geneva Convention (IV) Relative to the Protection of Civilian Persons in Time of War (adopted 12

August 1949) 75 UNTS 135. 6 See, e.g., Protocol Relating to the Status of Refugees (adopted 31 January 1967) 606 UNTS 267.

7 See infra notes 8-15 and accompanying text.

Electronic copy available at: http://ssrn.com/abstract=1826102Electronic copy available at: http://ssrn.com/abstract=1826102

2

The functional analysis in Parts Two to Four shows that domestic courts play a key role

in protecting private rights under transnational treaty provisions and promoting compliance with

those provisions, but they play virtually no role in promoting compliance with horizontal treaty

provisions. This is generally true for both monist and dualist states. The story with respect to

vertical treaty provisions is more complicated. When domestic courts adopt a transnationalist

approach, they play a key role in protecting private rights under vertical treaty provisions and

promoting compliance with those provisions. When domestic courts adopt a nationalist

approach, vertical treaty provisions may be under-enforced. There does not appear to be any

significant correlation between a state‘s formal classification as monist or dualist and the

tendency of domestic courts in that state to function in a nationalist or transnationalist mode.

I.

Monism and Dualism

The terms ‗monism‘ and ‗dualism‘ generate considerable confusion because there is no

single, agreed definition of the terms. Some scholars employ the terms to describe contrasting

theoretical perspectives on the relationship between international and domestic law.8 Used in this

sense, dualism ‗points to the essential difference of international law and municipal law,

consisting primarily in the fact that the two systems regulate different subject-matter‘.9 In

contrast, monism holds that ‗international and municipal law are part of the same system of

norms‘.10

Some monist theorists assert ‗the supremacy of international law‘ over domestic law,

but this is not an essential feature of monist theory.11

Other scholars employ the terms monism and dualism to describe different types of

domestic legal systems.12

Used in this sense, dualist states are states in which ‗the constitution ...

accords no special status to treaties; the rights and obligations created by them have no effect in

domestic law unless legislation is in force to give effect to them‘.13

In contrast, ‗[t]he essence of

the monist approach is that a treaty may, without legislation, become part of domestic law once it

has been concluded in accordance with the constitution and has entered into force for the state‘.14

As Professor Aust correctly notes, many national constitutions ‗contain both dualist and monist

elements‘.15

This chapter uses the terms monism and dualism in the second sense, to describe different

types of domestic legal systems. Dualist states are states in which no treaties have the status of

law in the domestic legal system; all treaties require implementing legislation to have domestic

8 See, e.g., Ian Brownlie, Principles of Public International Law (7

th edn OUP, Oxford 2008) 31–33.

9 ibid 31.

10 ibid 32.

11 See ibid 32–33 (discussing Kelsen‘s and Lauterpacht‘s theories).

12 See, e.g., Anthony Aust, Modern Treaty Law and Practice (CUP, Cambridge 2007) 181–95.

13 ibid 187.

14 ibid 183.

15 ibid 182.

3

legal force.16

Monist states are states in which some treaties have the status of law in the

domestic legal system, even in the absence of implementing legislation.17

In most monist states,

there are some treaties that require implementing legislation and others that do not. There is

substantial variation among monist states as to which treaties require implementing legislation.

Moreover, monist states differ considerably in terms of the hierarchical rank of treaties within

the domestic legal order. Despite these variations, all monist states have one common feature: at

least some treaties have the status of law within the domestic legal order.

The question whether a treaty requires legislative implementation after the treaty enters

into force internationally must be distinguished from the question whether legislative approval is

necessary prior to treaty ratification. In most dualist states, the executive has the constitutional

authority to conclude treaties that bind the nation under international law without obtaining prior

legislative approval.18

The executive‘s power to conclude treaties without prior legislative

approval helps explain why, in dualist states, implementing legislation is necessary to grant

treaties domestic legal force. In most monist states, though, the constitution requires legislative

approval for at least some treaties before the executive can make an internationally binding

commitment on behalf of the nation.19

The fact that the legislature approves (some) treaties

before they become binding on the nation helps explain why, in monist states, some treaties have

the status of domestic law even in the absence of implementing legislation. In sum, in both

monist and dualist states, it is rare for a treaty to have domestic legal force unless the legislature

has acted either to approve the treaty before international entry into force, or to implement the

treaty after international entry into force.20

The following sections summarize key features of monist and dualist systems. The

analysis touches upon the domestic legal systems of twenty-one states, relying heavily on two

previously published volumes that present a comparative analysis of national treaty law.21

Those

twenty-one states include five dualist states: Australia, Canada, India, Israel and the United

Kingdom. The other sixteen (monist) states are: Austria, Chile, China, Columbia, Egypt, France,

Germany, Japan, Mexico, Netherlands, Poland, Russia, South Africa, Switzerland, Thailand, and

the United States.

A. Dualist States

Almost all the British Commonwealth states follow the dualist approach for treaties.22

Apart from Commonwealth states, Israel, Denmark and other Nordic states also follow a dualist

16

In many dualist states, customary international law has domestic legal force, even in the absence of implementing

legislation. See, e.g., Nihal Jayawickrama, ‗India‘ in David Sloss (ed), The Role of Domestic Courts in Treaty

Enforcement: A Comparative Study (CUP, Cambridge 2009) 244–45. 17

These definitions arguably constitute a slight departure from standard terminology. However, these definitions

have the advantage of drawing a clear distinction between monism and dualism. Applying these definitions, almost

all states can be neatly classified as either monist or dualist, without any significant overlap between the categories. 18

See infra note 25. 19

See infra note 72 and accompanying text. 20

See Duncan B. Hollis, ‗A Comparative Approach to Treaty Law and Practice‘ in Duncan B. Hollis, Merritt R.

Blakeslee & L. Benjamin Ederington (eds), National Treaty Law and Practice (Martinus Nijhoff, Leiden, Boston

2005) 32–45 [hereinafter National Treaty Law]. 21

See National Treaty Law (n 20); Sloss (n 16). 22

See Aust (n 12) 194–95.

4

approach.23

The key distinguishing feature of dualism is that no treaties have the formal status of

law in the domestic legal system unless the legislature enacts a statute to incorporate the treaty

into domestic law.24

Such statutes must be distinguished from legislative acts that authorize the

executive to make a binding international commitment. In dualist states, the executive typically

has the constitutional authority to make a binding international commitment on behalf of the

nation without obtaining prior legislative approval.25

However, in many dualist states the

executive consults with the legislature before concluding ‗important‘ treaties.26

(There is

considerable variation among states concerning which treaties qualify as ‗important‘.) Moreover,

if legislation is needed to ensure that government officials have the requisite authority to

implement a treaty, dualist states usually enact the necessary implementing legislation before the

treaty enters into force internationally.27

For courts in dualist states, there is a crucial distinction between incorporated and

unincorporated treaties. As a formal matter, courts in dualist states have no authority to apply

treaties directly as law. If the legislature has enacted a statute to incorporate a particular treaty

provision into national law, courts apply the statute as law;28

they frequently consult the

underlying treaty to help construe the meaning of the statute.29

Thus, in dualist states, courts

apply treaties indirectly, not directly. However, one should not overstate the difference between

direct and indirect application. In practice, courts can achieve roughly the same results, whether

they apply the treaty directly or indirectly. Either way, judges who are receptive to the domestic

judicial application of treaties can use their judicial power to protect the treaty-based rights of

private parties and promote compliance with national treaty obligations.30

Dualist states employ a variety of methods for incorporating treaties into national law.31

In the United Kingdom, for example: the text of a treaty may be attached to a statute stipulating

that the attached treaty provisions ‗shall have the force of law in the United Kingdom‘;32

Parliament may pass an Act granting government officials ‗all the powers necessary to carry out

23

See ibid. 24

See Donald R. Rothwell, ‗Australia‘ in Sloss (n 16) 128–30; Maurice Copithorne, ‗National Treaty Law and

Practice: Canada‘ in National Treaty Law (n 20) 95–101; Dr. K. Thakore, ‗National Treaty Law and Practice: India‘

in National Treaty Law (n 20) 351; Ruth Lapidoth, ‗National Treaty Law and Practice: Israel‘ in National Treaty

Law (n 20) 396; and Sir Ian Sinclair, Susan J. Dickson and Graham Maciver, ‗National Treaty Law and Practice:

United Kingdom‘ in National Treaty Law (n 20) 733. 25

See Copithorne (n 24) 91–94 (Canada); Lapidoth (n 24) 385–90 (Israel); Rothwell (n 24) 128–30 (Australia);

Sinclair and others (n 24) 727 (United Kingdom); and Thakore (n 24) 352–55 (India). 26

See Copithorne (n 24) 96, 98 (Canada); Lapidoth (n 24) 388–89, 393–94 (Israel); Sinclair and others (n 24) 737–

39 (United Kingdom); and Thakore (n 24) 365–66 (India). 27

See Copithorne (n 24) 96 (Canada); Lapidoth (n 24) 396–98 (Israel); Sinclair and others (n 24) 742 (United

Kingdom); and Thakore (n 24) 359–60 (India). 28

See, e.g., Anthony Aust, ‗United Kingdom‘ in Sloss (n 16) 486; Rothwell (n 24) 138–41 (Australia); and Gib van

Ert, ‗Canada‘ in Sloss (n 16) 202–04. 29

See, e.g., Aust (n 28) 482–83 (United Kingdom); Jayawickrama (n 16) 264–66 (India); David Kretzmer, ‗Israel‘

in Sloss, (n 16) 290–92 (Israel); Rothwell (n 24) 138–41 (Australia); and van Ert (n 28) 175–82 (Canada). 30

See generally David Sloss, ‗Treaty Enforcement in Domestic Courts: A Comparative Analysis‘ in Sloss (n 16) 8–

43 (analyzing the practice of national courts in eleven states). 31

See, e.g., Kretzmer (n 29) 283–85 (Israel); Rothwell (n 24) 159–60 (Australia); and van Ert (n 28) 169–71

(Canada). 32

Aust (n 12) 189.

5

obligations under an existing or future treaties‘;33

or Parliament may pass an Act authorizing the

Crown to enact regulations to implement one or more treaties.34

Given the wide variety of

techniques that dualist states utilize to incorporate treaties,35

the question whether a particular

treaty provision has been incorporated is often ambiguous.36

The Australian High Court developed a creative approach to addressing this type of

ambiguous situation, which commentators have dubbed ‗quasi-incorporation‘.37

The term refers

to situations where ‗government departments, and administrative decision makers are given [a

statutory directive] to take into account the provisions of ... international instruments to which

Australia is a party‘.38

For example, in the Project Blue Sky case,39

an Australian statute

specifically directed the Australian Broadcasting Authority (ABA) ‗to perform its functions in a

manner consistent with ―Australia‘s obligations under any ... agreement between Australia and a

foreign country‖‘.40

The petitioners argued that the ABA had violated the statute by enacting

regulations inconsistent with a bilateral free-trade agreement between Australia and New

Zealand.41

A three-judge panel of the Federal Court held that ‗the ABA was not bound to take

into account‘ the free-trade agreement because that agreement conflicted with a different

statutory provision.42

The High Court reversed, holding ‗that the ABA was precluded from

making a standard inconsistent with the‘ free-trade agreement, even though that agreement had

not been directly incorporated into Australian domestic law.43

Courts in other dualist states have

adopted a similar approach. In the United Kingdom, for example, petitioners in several cases

have obtained judicial remedies by invoking statutes that required administrative decision

makers to exercise their authority in conformity with treaty obligations that had not been directly

incorporated into domestic law.44

More surprisingly, courts in dualist states have developed a variety of strategies for

judicial application of unincorporated treaties — even in the absence of any statutory directive

33

ibid 190. 34

ibid 190–91. 35

See, e.g., Aust (n 28) 479–81 (United Kingdom) (discussing, among others, Cheng v Conn, Inspector of Taxes

[1968] 1 All ER 779); Kretzmer (n 29) 283–85 (Israel); Rothwell (n 24) 158–60 (Australia) (discussing Project Blue

Sky Inc v Australian Broadcasting Auth (1998) 153 ALR 490); and van Ert (n 28) 169–71 (Canada) (discussing,

among othters, Pan American World Airways v The Queen [1981] 2 SCR 565; Schavernoch v Foreign Claims

Commission [1982] 1 SCR 1092). 36

See, e.g., van Ert (n 28) 171 (stating ‗that the absence of formal rules on how treaties are implemented can create

uncertainty about whether treaties have been implemented at all‘). 37

See Rothwell (n 24) 158–64. 38

ibid 159. 39

Project Blue Sky Inc. v Australian Broadcasting Auth. (1998) 153 ALR 490. 40

Rothwell (n 24) 141 (quoting Broadcasting Services Act 1992). 41

ibid 141–42. 42

ibid 143. 43

ibid 143–45. 44

See Aust (n 28) 490–91 (noting that ‗there have been numerous successful challenges by way of judicial review to

[administrative] decisions on claims to refugee status‘); ibid 491–92 (discussing Secretary of State for Foreign &

Commonwealth Affairs v Quark Fishing Ltd [2002] EWCA Civ 1409; [2002] AER (D) 450 holding that the Director

of Fisheries of South Georgia and the South Sandwich Islands ‗had not properly carried out his statutory powers‘

because he failed to take account of relevant treaty provisions).

6

for government officials to take account of treaty provisions.45

In Australia, for example, the

High Court held in Minister of State for Immigration and Ethnic Affairs v Teoh46

that

administrative decision makers must exercise their statutory discretion in conformity with the

Convention on the Rights of the Child, an unincorporated treaty, because treaty ratification

meant that individuals had a ‗legitimate expectation‘ that government officials would act in

accordance with the treaty.47

The Canadian Supreme Court has declined to follow this so-called

legitimate expectations doctrine.48

Even so, the Canadian Supreme Court has held that

administrative decision makers in Canada, like their Australian counterparts, must exercise their

statutory discretion in conformity with the Convention on the Rights of the Child, an

unincorporated treaty.49

In Israel, ‗it has now become standard practice for the Supreme Court to‘

apply Geneva Convention IV in cases involving the Occupied Territories, although the

Convention has not been incorporated into domestic law.50

The Court justifies this approach by

citing the government‘s political commitment to ‗respect the humanitarian provisions of the

Convention‘.51

Similarly, the Indian Supreme Court routinely applies unincorporated treaties to

support its interpretation of both statutory and constitutional provisions;52

the Court has also

applied treaties to support its progressive development of common law principles.53

This increasing judicial reliance on unincorporated treaties by courts in dualist states

blurs the traditional distinction between monist and dualist states.54

Nevertheless, judges in

dualist states periodically invoke the dualist dogma that courts are powerless to apply treaties

unless the legislature has expressly incorporated the treaty into domestic law.55

Hence, there

remains an uneasy tension between the formalities of strict dualist doctrine and the practical

reality that courts in dualist states have developed a variety of strategies to facilitate judicial

application of unincorporated and partially incorporated treaties.

B. Monist States

45

See Michael P. Van Alstine, ‗The Role of Domestic Courts in Treaty Enforcement: Summary and Conclusions‘ in

Sloss (n 16) 608–12. 46

(1995) 128 ALR 353. 47

See Rothwell (n 24) 146–48 (quoting Minister of State for Immigration and Ethnic Affairs v Teoh (1995) 128

ALR 353). 48

See van Ert (n 28) 173 (discussing Baker v Canada (1999) 2 SCR 817). 49

See ibid 194–97 (discussing Baker v Canada (1999) 2 SCR 817). 50

See Kretzmer (n 29) 305–10 (discussing, among others, HCJ 3278/02, Hamoked The Center for the Defense of the

Individual v IDF Commander 57 P.D. (1) 385). 51

ibid 309–10 (discussing HCJ 3278/02, Hamoked The Center for the Defense of the Individual v IDF Commander

57 P.D. (1) 385; HCJ 7862/04, Abu Dahar v IDF Commander 59 P.D. (5) 368). 52

See Jayawickrama (n 16) 247–64 (discussing, among others, Jolly George Verhese v Bank of Cochin (1980) 2

SCR 913; Transmission Coprporation of Andhra Pradesh v Ch. Prabhakar Supreme Court of India, Civil Appeal

6131 of 2002, 26 May 2004). 53

See ibid 255–56 (discussing MV Elisabeth v Harwan Investment and Trading Pvt Ltd [1992] 1 SCR 1003). 54

See Melissa A. Waters, ‗Creeping Monism: The Judicial Trend Toward Interpretive Incorporation of Human

Rights Treaties‘ (2007) 107 Colum. L. Rev. 628. 55

See, e.g., Ben Saul, ‗The Kafka-esque Case of Sheikh Manour Leghaei: The Denial of the International Human

Right to a Fair Hearing in National Security Assessments and Migration Proceedings in Australia‘ (2010) Sydney

Law School, Legal Studies Research Paper No. 10/111 <http://ssrn.com/abstract=1701374> accessed 27 March

2011.

7

The key distinguishing feature of monist legal systems, as defined herein, is that at least

some treaties are incorporated into the domestic legal order without the need for any legislative

act, other than the act authorizing the executive to conclude the treaty. Under this definition,

Austria,56

Chile,57

China,58

Columbia,59

Egypt,60

France,61

Germany,62

Japan,63

Mexico,64

the

Netherlands,65

Poland,66

Russia,67

South Africa,68

Switzerland,69

Thailand,70

and the United

States71

all have monist legal systems. In all sixteen states, some form of legislative approval is

required for at least some types of treaties before the executive is authorized to make a binding

international commitment on behalf of the nation.72

Despite these similarities, there are

substantial differences among these states concerning the application of treaties within their

national legal systems.

One significant area of variability concerns the types of treaties that require legislative

approval before international entry into force of the treaty.73

In Mexico and Colombia, all treaties

require prior legislative approval.74

Chile, the Netherlands, South Africa, and Switzerland

establish a default rule that treaties ordinarily require legislative approval, but they recognize

certain exceptions to that rule.75

In other states, legislative approval is required only for

designated categories of treaties.76

56

See Franz Cede & Gerhard Hafner, ‗National Treaty Law and Practice: Republic of Austria‘ in National Treaty

Law (n 20) 59–60, 67–68. 57

See Francisco Orrego Vicuna & Francisco Orrego Bauzá, ‗National Treaty Law and Practice: Chile‘ in National

Treaty Law (n 20) 136–38. 58

See Xue Hanqin, Hu Zhiqiang & Fan Kun, ‗National Treaty Law and Practice: China‘ in National Treaty Law (n

20) 163–64. 59

See Germán Cavelier, ‗National Treaty Law and Practice: Colombia‘ in National Treaty Law (n 20) 205. 60

See Nabil Elaraby, Mohammed Gomaa, & Lamia Mekhemar, ‗National Treaty Law and Practice: Egypt‘ in

National Treaty Law (n 20) 238–39. 61

See Pierre Michel Eisemann & Raphaële Rivier, ‗National Treaty Law and Practice: France‘ in National Treaty

Law (n 20) 265–67. 62

See Dr. Hubert Beemelmans & Dr. Hans D. Treviranus, ‗National Treaty Law and Practice: Federal Republic of

Germany‘ in National Treaty Law (n 20) 323–26. 63

See Takao Kawakami, ‗National Treaty Law and Practice: Japan‘ in National Treaty Law (n 20) 424–25. 64

See Dr. Luis Miguel Díaz, ‗National Treaty Law and Practice: Mexico‘ in National Treaty Law (n 20) 451. 65

See J.G. Brouwer, ‗National Treaty Law and Practice: The Netherlands‘ in National Treaty Law (n 20) 497–99. 66

See Lech Garlicki, Malgorzata Masternak-Kubiak, & Krzysztof Wójtowicz, ‗Poland‘ in Sloss (n 16) 378. 67

See W.E. Butler, ‗National Treaty Law and Practice: Russia‘ in National Treaty Law (n 20) 554–56. 68

See N.J. Botha, ‗National Treaty Law and Practice: South Africa‘ in National Treaty Law (n 20) 600–02. 69

See Luzius Wildhaber, Adrian Scheidegger, & Marc D. Schinzel, ‗National Treaty Law and Practice: Switzerland‘

in National Treaty Law (n 20) 658–59. 70

See Sompong Sucharitkul, ‗National Treaty Law and Practice: Thailand‘ in National Treaty Law (n 20) 706. 71

See Robert E. Dalton, ‗National Treaty Law and Practice: United States‘ in National Treaty Law (n 20) 788–90. 72

See Beemelmans & Treviranus (n 62) 323–26 (Germany); Botha (n 68) 590–92 (South Africa); Brouwer (n 65)

489–91 (the Netherlands); Butler (n 67) 544–47 (Russia); Cavelier (n 59) 199 (Colombia); Cede & Hafner (n 56)

64–65 (Austria); Dalton (n 71) 770–74 (United States); Díaz (n 64) 447–48 (Mexico); Eisemann & Rivier (n 61)

258–60 (France); Elaraby and others (n 60) 231 (Egypt); Garlicki and others (n 66) 376–77 (Poland); Hanqin and

others (n 58) 161–62 (China); Kawakami (n 63) 419–20 (Japan); Sucharitkul (n 70) 701–03 (Thailand); Vicuna &

Bauzá (n 57) 127–30 (Chile); and Wildhaber and others (n 69) 644–48 (Switzerland). 73

For a tabular depiction of the variability in this area, see Hollis ‗Comparative Approach‘ (n 20) 33. 74

See Cavelier (n 59) 199 (Colombia); Díaz (n 64) 447–48 (Mexico). 75

See Botha (n 68) 586–92 (South Africa); Brouwer (n 65) 489–91 (the Netherlands); Vicuna & Bauzá (n 57) 123–

24 (Chile); and Wildhaber and others (n 69) 644–51 (Switzerland). 76

See Hollis, ‗Comparative Approach‘ (n 20) 32–37.

8

Another significant area of variability relates to publication requirements. In Egypt,

France, Chile, Japan, and Russia, a treaty that has entered into force internationally lacks

domestic legal force until the executive branch publishes or promulgates the treaty

domestically.77

In other monist states, though, (at least some) treaties enter into force

domestically at the same time they enter into force internationally, without the need for any

additional steps.78

There is also significant variation among monist states concerning the hierarchical rank

of treaties within the domestic legal order. In Austria, Egypt, Germany, and the United States,

treaties are equivalent to statutes; they rank lower than the Constitution.79

In South Africa,

treaties rank lower than statutes.80

In China, France, Japan, Mexico, and Poland, (at least some)

treaties rank higher than statutes but lower than the Constitution.81

In the Netherlands, some

treaties rank higher than the Constitution.82

In Chile, Russia and Switzerland, the hierarchical

rank of treaties is contested, but it is undisputed that at least some treaties rank higher than

statutes,83

and there is some authority for the proposition that some treaties have constitutional

rank.84

In many monist states, even if a treaty has the formal status of law in the absence of

implementing legislation, the legislature sometimes enacts legislation to help ensure that courts

and executive officers give practical effect to the treaty within the national legal system. Thus,

for example, the United States enacted implementing legislation for the New York Convention,85

and South Africa enacted implementing legislation for the Warsaw Convention.86

As Professor

77

See Butler (n 67) 552–54 (Russia); Eisemann & Rivier (n 61) 265–67 (France); Elaraby and others (n 60) 238–39

(Egypt); Kawakami (n 63) 424–25 (Japan); Vicuna & Bauzá (n 57) 136–38 (Chile). 78

See Hollis, ‗Comparative Approach‘ (n 20) 41–42. 79

See Cede & Hafner (n 56) 59–60, 67–68 (Austria); Dalton (n 71) 789–90 (United States); Elaraby and others (n

60) 238–39 (Egypt); Andreas L. Paulus, ‗Germany‘ in Sloss (n 16) 214–18. In both Austria and Germany, treaties

approved by the legislature have the rank of statutes, but treaties concluded without legislative approval have a

lower rank. See Cede & Hafner (n 56) 67–68; Paulus, ‗Germany‘ in Sloss (n 16) 214–18. In the United States,

though, there is at least some authority for the proposition that treaties concluded without legislative approval have

the same rank as treaties approved by the legislature. See United States v Pink 315 US 203, 62 S Ct 552 (1942);

United States v Belmont 301 US 324, 57 S Ct 758 (1937). 80

This follows directly from Article 231(4) of the South African Constitution, which states: ‗Any international

agreement becomes law in the Republic when it is enacted into law by national legislation; but a self-executing

provision of an agreement that has been approved by Parliament is law in the Republic unless it is inconsistent with

the Constitution or an Act of Parliament‘. 81

See Díaz (n 64) 451–54 (Mexico); Eisemann & Rivier (n 61) 263–67 (France); Garlicki and others (n 66) 376–79

(Poland); Hanqin and others (n 58) 163–65 (China); and Timothy Webster, ‗International Human Rights Law in

Japan: The View at Thirty‘ (2010) 23 Colum. J. Asian L. 241, 245. 82

See Brouwer (n 65) 498–99. 83

See Butler (n 67) 554–56 (Russia); Vicuna & Bauzá (n 57) 138–39 (Chile); and Wildhaber and others (n 69) 658–

64 (Switzerland). 84

See Butler (n 67) 556 (contending that ‗[t]he primacy of international treaties of the Russian Federation extends to

Federal laws, including constitutional laws‘.); Vicuna & Bauzá (n 57) 139 (noting that, in one view, human rights

treaties ‗now have in Chile a ranking above that of ordinary statutes and at least equal to the Constitution‘); and

Wildhaber and others (n 69) 662 (Switzerland) (‗Treaties in conflict with federal constitutional law have to be

applied irrespective of their unconstitutionality‘.). 85

See Federal Arbitration Act 1970 ss 201–08 (implementing the New York Convention). 86

See John Dugard, ‗South Africa‘ in Sloss (n 16) 470.

9

Nollkaemper observes: ‗[E]ven if the provisions of a treaty could in principle be applied directly,

the Netherlands usually chooses to convert them into national legislation to harmonize Dutch law

with the requirements of international law‘.87

All monist states recognize the possibility, at least theoretically,88

that domestic courts

can apply (at least some) treaties directly as law.89

Indeed, this is one of the crucial differences

between monist and dualist systems: dualist states permit only indirect judicial application of

treaties, whereas monist states permit direct judicial application in some cases. Despite this

formal distinction, however, there are several reasons why judicial practice exhibits many

similarities between monist and dualist states. First, as noted above, courts in dualist states apply

various strategies to facilitate judicial application of unincorporated and partially incorporated

treaties.90

Second, courts in monist states often apply treaties indirectly as an aid to statutory or

constitutional interpretation, rather than applying treaties directly as rules of decision to resolve

disputed issues.91

It is difficult to measure the relative frequency of direct versus indirect

application, but there is some evidence that courts even in monist states rely more heavily on

indirect than direct application.92

Indeed, courts may prefer indirect application ‗in cases where

the direct application of international law would conflict with national law‘ because ‗[c]ourts

usually prefer a conciliatory solution over the acknowledgment and resolution of a conflict of

law‘.93

Insofar as courts in monist states prefer indirect rather than direct application, this further

erodes the practical significance of the traditional distinction between monist and dualist states.

Finally, in certain monist states, courts have articulated a distinction between ―self-

executing‖ and ―non-self-executing‖ treaties.94

When domestic courts decide that a treaty is

―non-self-executing,‖ they sometimes behave as if the treaty has not been incorporated into

domestic law even though the treaty, as a formal matter, has the status of law within the domestic

legal system.95

Thus, just as judicial practice in some dualist states blurs the monist-dualist

divide by applying unincorporated treaties as if they were incorporated, judicial practice in some

87

André Nollkaemper, ‗The Netherlands‘ in Sloss (n 16) 335. 88

The South African Constitution expressly contemplates a category of self-executing treaties, but South African

courts have not yet held that any particular treaty is self-executing. See Dugard (n 86) 453–55. 89

See William E. Butler, ‗Russia‘ in Sloss (n 16) 410–11; Cede & Hafner (n 56) 69 (Austria); Dalton (n 71) 788–90

(United States); Díaz (n 64) 454 (Mexico); Eisemann & Rivier (n 61) 265–70 (France); Elaraby and others (n 60)

238–39 (Egypt); Garlicki and others (n 66) 400–04 (Poland); Hanqin and others (n 58) 163–65 (China);

Nollkaemper (n 87) 341–48 (the Netherlands); Paulus (n 79) 209–12 (Germany); Vicuna & Bauzá (n 57) 136–39

(Chile); Webster (n 81) 244–47 (Japan); and Wildhaber and others (n 69) 644–48 (Switzerland). 90

See supra notes 37-53 and accompanying text. 91

See, e.g., Dugard (n 86) 457–63 (South Africa); Garlicki and others (n 66) 403–04 (Poland); Nollkaemper (n 87)

348–51 (the Netherlands); Paulus (n 79) 209–10 (Germany); and David Sloss, ‗United States‘ in Sloss (n 16) 526–

27. 92

See, e.g., Garlicki and others (n 66) 404 (stating that ‗the most typical technique [in Poland] is that of

coapplication of an international norm and a domestic norm‘). 93

Nollkaemper (n 87) 349. 94

See infra notes 161-77 and accompanying text. 95

In the United States, for example, courts behave as if non-self-executing treaties are unincorporated, even though

the Constitution states expressly that ―all treaties‖ are ―the supreme Law of the Land.‖ See Sloss, ‗United States‘ (n

91) 509–14, 527-29, 534-39.

10

monist states blurs the monist-dualist divide by handling formally incorporated treaties as if they

were unincorporated.

II.

Horizontal, Transnational and Vertical Treaty Provisions

To appreciate the role of domestic courts in treaty application, it is important to

understand the nature of modern treaties. There is a widespread misconception that treaties focus

exclusively, or almost exclusively, on regulating horizontal relations among states. This was

never really true,96

and it is certainly not true in the twenty-first century. States conclude treaties

to regulate three different types of relationships: horizontal relations between and among states,

vertical relations between states and private actors (including natural persons and corporations),

and transnational relations between private actors who interact across national boundaries.97

The

role of domestic courts in applying treaties varies greatly depending on whether the treaty

provision at issue is horizontal, vertical or transnational.98

Domestic courts rarely apply treaties that regulate horizontal relationships among states.

If one state believes that another state has violated a horizontal treaty obligation, the complainant

might raise the issue in diplomatic negotiations, or perhaps file suit in an international tribunal,

but it would be unusual for the complainant to file suit in a domestic court. Domestic courts

typically dismiss cases in which private litigants file suit to resolve disputes that are properly

characterized as horizontal disputes between states, because domestic courts generally lack the

institutional competence to adjudicate such disputes. For example, a group of Serbian citizens

sued the Dutch government in a domestic court in the Netherlands, alleging that the government

violated Article 2(4) of the U.N. Charter by supporting the NATO bombing of Yugoslavia in

1999. The Supreme Court of the Netherlands held that plaintiffs were not entitled to invoke

Article 2(4) in a Dutch court.99

U.S. Supreme Court Chief Justice Marshall made a similar point

two centuries ago. Speaking as a Member of Congress (before he was appointed to the Supreme

Court), he asserted that a treaty-related claim falls within the scope of judicial competence where

parties ‗come into court, who can be reached by its process, and bound by its power ... to which

they are bound to submit‘.100

However, in a case where ‗[t]he parties were the two nations ... the

demand is not a case for judicial cognizance‘101

because sovereign nations are generally not

bound to submit to the power of domestic courts.

In contrast to horizontal treaties, domestic courts routinely apply transnational treaty

provisions that regulate cross-border relationships between private actors. Such treaties include,

96

See, e.g., David Sloss, ‗When Do Treaties Create Individually Enforceable Rights?‘ (2006) 45 Colum. J. Trans‘l

L. 20, 51–91 (analyzing U.S. Supreme Court‘s application of vertical and transnational treaty provisions between

1789 and 1838). 97

A separate category of treaties involves agreements between States and international organizations. Such treaties

involve horizontal provisions (such as a nation‘s obligation to make financial contributions) and vertical provisions

(such as immunities for employees of international organizations). Treaties between states and international

organizations do not generally include transnational provisions. 98

In assessing whether a particular treaty provision is properly characterized as horizontal, vertical, or transnational,

it is important to examine the specific provision at issue because a single treaty may contain a combination of

horizontal, vertical and transnational provisions. 99

See Nollkaemper (n 87) 347. 100

10 Annals of Cong. 613 (1800). 101

ibid.

11

for example, the 1958 New York Convention,102

the 1999 Montreal Convention,103

and the 1980

Hague Convention on Child Abduction.104

Although states negotiated and ratified these treaties,

they are designed primarily to regulate cross-border relationships among private actors, not

horizontal relationships among states. The New York Convention provides rules for recognition

and enforcement of arbitral awards arising from transnational commercial activities. The

Montreal Convention governs relationships between airlines and their customers: both

passengers and shippers. The Hague Convention applies to child custody disputes in which one

parent transports a child across national boundaries. For these and other transnational treaties,

domestic courts play a vital role in ensuring that private actors behave in accordance with

internationally agreed rules regulating cross-border activities. Indeed, domestic courts are

arguably the primary enforcers of transnational treaty obligations because most international

tribunals lack jurisdiction to adjudicate private disputes involving alleged infractions of

transnational treaty provisions.105

Moreover, such disputes rarely have sufficient political

salience to become the subject of interstate diplomacy.

The preceding comments apply equally to monist and dualist states. Although there are

significant formal distinctions between monist and dualist states (as discussed in Part One

above), there are few, if any, functional distinctions. In both monist and dualist states, domestic

courts rarely apply horizontal treaty provisions, but they routinely apply transnational treaty

provisions.

The most significant differences among states relate to the judicial application of vertical

treaty provisions — provisions that regulate relations between states and private parties.

Prominent examples of vertical treaty provisions include the Covenant on Civil and Political

Rights (which protects the civil and political rights of citizens in relation to their own

governments)106

and the Refugee Protocol (which protects the rights of individuals who have

fled persecution in their home countries to seek asylum in other countries).107

Domestic courts in

both monist and dualist states apply vertical treaty provisions more frequently than they apply

horizontal treaty provisions because, in most mature legal systems, domestic courts have an

institutional responsibility to protect the rights of private parties, and vertical treaties (unlike

horizontal treaties) create rights for private parties.

Whereas both vertical and transnational treaty provisions implicate the rights of private

parties — and therefore invite judicial application of treaties — vertical treaty provisions

implicate the public functions of government in a way that is not true for transnational treaty

102

Convention on the Recognition and Enforcement of Foreign Arbitral Awards (adopted 10 June 1958) 330 UNTS

38. 103

Convention for the Unification of Certain Rules for International Carriage by Air (adopted 28 May 1999) 2242

UNTS 309. 104

Convention on the Civil Aspects of International Child Abduction (adopted 25 October 1980) 1343 UNTS 89. 105

Arbitral panels established pursuant to bilateral investment treaties frequently adjudicate disputes between states

and private corporations but – in contrast to international commercial arbitration – investment treaty arbitration

typically involves vertical treaty provisions, not transnational treaty provisions. The International Court of Justice

occasionally adjudicates disputes that originated as transnational, commercial disputes between private parties. See,

e.g., Jurisdiction and Enforcement of Judgments in Civil and Commercial Matters (Belgium v Switzerland) ICJ

Press Release 2009/36. However, these types of transnational, private disputes rarely give rise to ICJ jurisdiction. 106

International Covenant on Civil and Political Rights (n 4). 107

Protocol Relating to the Status of Refugees (n 6).

12

provisions. For example, the Refugee Protocol regulates the public functions of government by

creating legal (vertical) duties that the government owes to individuals who claim refugee status

under the treaty. In contrast, the 1999 Montreal Convention108

regulates the cross-border

commercial activities of airlines, including state-owned airlines, but it does not create significant

new duties for governments in the exercise of traditional public functions.

This distinction between vertical and transnational treaty provisions helps explain the

distinction between nationalist and transnationalist approaches to the judicial application of

treaties. ‗Transnationalist‘ decisions manifest a belief that the judiciary has an independent

responsibility to ensure that domestic government officials act in accordance with international

treaty obligations. ‗Nationalist‘ decisions manifest a belief that courts should not scrutinize too

closely government conduct that is arguably inconsistent with international treaty obligations. In

countries where courts adopt a more ―transnationalist‖ approach — such as South Africa109

and

the Netherlands110

— domestic courts apply both vertical and transnational treaty provisions with

equal vigor. However, in states where courts adopt a more ―nationalist‖ approach — such as the

United States111

and Israel112

— domestic courts are hesitant to apply vertical treaty provisions,

even though they routinely apply transnational provisions.113

The contrast between nationalist and transnationalist approaches manifests different

judicial attitudes about the relative weight assigned to two competing factors: the judicial

responsibility to protect the rights of private parties and the judicial responsibility to refrain from

interfering with public governmental functions.114

Transnationalist judges assign greater weight

(implicitly, if not explicitly) to the judicial responsibility to protect the rights of private parties,

including rights vis-à-vis government actors protected by vertical treaty provisions. Nationalist

judges assign greater weight (again implicitly, if not explicitly) to the judicial responsibility to

defer to the political branches‘ judgment about how best to interpret and apply vertical treaty

provisions. It bears emphasis that the distinction between nationalist and transnationalist

approaches is best conceptualized as a spectrum with multiple shades of gray, not a sharp line

separating black and white.

One might think that courts in monist states are more transnationalist and courts in dualist

states are more nationalist. In fact, though, there is not any significant correlation along these

lines. Courts in dualist states sometimes adopt a transnationalist approach and courts in monist

states sometimes adopt a nationalist approach.115

Hence, the monist-dualist dichotomy cannot

108

Convention for the Unification of Certain Rules for International Carriage by Air (n 102). 109

See Dugard (n 86) 448–75. 110

See Nollkaemper (n 87) 326–69. 111

See Sloss, ‗United States‘ (n 91) 504–54. 112

See Kretzmer (n 29) 273–325. 113

For more detailed analysis, see Sloss, ‗Treaty Enforcement‘ (n 30) 1–60; see also Van Alstine (n 45) 555–613. 114

See Paul B. Stephan, ‗Treaties in the Supreme Court, 1946–2000‘ in David L. Sloss, Michael D. Ramsey, and

William S. Dodge (eds), International Law in the U.S. Supreme Court: Continuity and Change (CUP, Cambridge

2011) [hereinafter Intl Law in the U.S. Supreme Court] (discussing U.S. Supreme Court‘s reluctance to apply

treaties in a manner that would constrain the executive branch in its exercise of public governmental functions). 115

No state is purely nationalist and no state is purely transnationalist. However, courts in some states have more

nationalist tendencies and courts in other states have more transnationalist tendencies. See Sloss, ‗Treaty

Enforcement‘ (n 30). To obtain more accurate information, a detailed, multi-state empirical study is needed. No such

13

explain variations among states in judicial decision-making in cases involving vertical treaty

provisions. Rather, the extent to which domestic courts apply vertical treaty provisions is best

explained by examining whether courts in a particular country are more inclined to adopt a

nationalist or transnationalist approach.

III.

Nationalist and Transnationalist Approaches

Part Three discusses nationalist and transnationalist techniques that courts apply,

focusing primarily on cases in which litigants ask courts to apply vertical treaty provisions. The

tension between nationalist and transnationalist approaches generally does not arise in cases

involving horizontal treaty provisions because courts rarely apply horizontal treaty provisions.

Similarly, the tension between nationalist and transnationalist approaches rarely arises in cases

involving transnational treaty provisions: courts in both monist and dualist states routinely apply

transnational treaty provisions without hesitation.

The fact that the tension between nationalist and transnationalist approaches pertains

primarily to vertical treaty provisions raises an additional point. Since vertical treaty provisions

regulate relations between states and private parties, litigated cases typically pit a private party

against a government actor. In some cases, the government invokes a vertical treaty provision to

support the exercise of governmental power to regulate private conduct.116

More commonly,

though, a private party invokes a vertical treaty provision as a constraint on government

action.117

Despite the spread of democratization since the end of the Cold War, many states still

lack a truly independent judiciary.118

In such states, transnationalism is not a viable option

because judges lack the institutional authority to issue and enforce judgments constraining

government conduct. In states that do have an independent judiciary, though, courts must still

decide whether to apply treaties — much as they would apply constitutional, statutory, or

common law — as a tool to constrain government action. Transnationalist judges apply treaties

in precisely this way, whereas nationalist judges employ various rationales for refraining to apply

treaties as a constraint on government action. This is the core feature of the distinction between

nationalist and transnationalist approaches.

study has been done, but the present author has done an empirical study of nationalist and transnationalist trends in

U.S. courts. See Sloss, ‗United States‘ (n 91). 116

For example, when the Security Council approved the transfer of Charles Taylor to the Netherlands to stand trial

before the Special Court for Sierra Leone, the government of the Netherlands relied on the Security Council

resolution, and therefore ultimately the U.N. Charter ‗to provide the proper legal basis in domestic law for the arrest

and detention of Charles Taylor‘. Nollkaemper (n 87) 329–30. 117

See, e.g., Paulus (n 79) 234–35 (discussing decisions of German courts applying Article 36 of the Vienna

Convention on Consular Relations) (discussing BGH, Beschluss v 7.11.2001 (Decision of 7 November 2001), 5 StR

116/01). 118

The Polity IV Project is the best source of data for estimating the number of countries with independent

judiciaries. The Polity IV Project rates 162 countries on a range of variables, one of which (xconst) ranks countries

in terms of constraints on the executive branch. See Polity IV Project, Political Regime Characteristics and

Transitions, 1800–2009: Dataset Users‘ Manual <http://www.systemicpeace.org/inscr/inscr.htm> accessed 26

March 2011. According to the most recent data, there are 81 countries that score 6 or 7 on the xconst variable,

meaning that there are significant constraints on the executive. This is a reasonably good proxy for determining

whether a country has an independent judiciary. Thus, approximately half the countries in the world have

independent judiciaries.

14

The following analysis of nationalist and transnationalist techniques is divided into four

sections: statutory interpretation, treaty interpretation, constitutional interpretation, and self-

execution. The first three sections address issues that are common to both monist and dualist

states. The final section addresses issues that are unique to monist states.119

A. Statutory Interpretation

Courts in both monist and dualist states frequently apply an interpretive presumption that

statutes should be construed in conformity with the nation‘s international legal obligations,

including obligations derived from both treaties and customary international law. This

interpretive presumption is sometimes called a ‗presumption of conformity‘ or a ‗presumption of

compatibility‘.120

In the United States, the presumption is referred to as the ‗Charming Betsy

canon‘.121

Labels aside, the presumption of conformity is probably the most widely used

transnationalist tool. Courts in Australia,122

Canada,123

Germany,124

India,125

Israel,126

the

Netherlands,127

Poland,128

South Africa,129

the United Kingdom,130

and the United States,131

among other countries, have applied the presumption in cases involving vertical treaty provisions

to help ensure that government conduct conforms to the nation‘s international treaty obligations.

One recurring issue concerns the threshold conditions necessary to trigger application of

the presumption. There is broad agreement that courts may apply the presumption in cases where

the statute is facially ambiguous. The Supreme Court of Canada has gone further, holding that ‗it

is reasonable to make reference to an international agreement at the very outset of the inquiry to

119

For a comparable analysis of issues unique to dualist states, see supra notes 37-55 and accompanying text. 120

See, e.g., van Ert (n 28) 188–97 (discussing application of the presumption of conformity by Canadian courts in

the context of, among others, R v Hape [2007] SCC 26); Kretzmer (n 29) 287–92 (discussing application of the

presumption of compatibility by Israeli courts in the context of, among others, Cr. A. 5/51, Steinberg v Attorney

General 5 P.D. 1061; HCJ 2599/00, Yated – Friendly Society of Downs Syndrome Children’s Parents v Ministry of

Education 56 P.D. (5) 834). 121

The canon takes its name from an 1804 decision by Chief Justice Marshall. See Murray v Schooner Charming

Betsy 6 US (2 Cranch) 64, 118 (1804). For an insightful analysis of the Supreme Court‘s application of the

Charming Betsy canon in the late twentieth century, see Melissa A. Waters, ‗International Law as an Interpretive

Tool in the Supreme Court, 1946–2000‘ in Intl Law in the U.S. Supreme Court (n 114). 122

See Rothwell (n 24) 152–56 (discussing, among others, Mabo v Queensland (No 2) (1992) 175 CLR 1). 123

See van Ert (n 28) 188–97 (discussing, among others, R v Hape [2007] SCC 26). 124

See Paulus (n 79) 209 (‗German courts are also bound to interpret domestic law, as far as possible, in a way that

avoids the breach of international legal obligations‘.) (citing BVerfGE 74, 358 at 370). 125

See Jayawickrama (n 16) 247–51 (discussing, among others, Jolly George Verghese [1980] 2 SCR 913). 126

See Kretzmer (n 29) 287–92 (discussing, among others, HCJ 2599/00, Yated – Friendly Society of Downs

Syndrom Children’s Parents v Ministry of Education 56 P.D. (5) 834). 127

See Nollkaemper (n 87) 348–51 (discussing, among others, Supreme Court, 27 May 2005, LJN AS7054). 128

See Garlicki and others (n 66) 404 (noting that ‗coapplication of an international norm and a domestic norm‘ is

the most common technique for the judicial application of treaties in Poland). 129

See Dugard (n 86) 457 (noting that the South African Constitution requires courts, when interpreting legislation,

to ‗prefer any reasonable interpretation of the legislation that is consistent with international law over any alternative

interpretation that is inconsistent with international law‘) (citing S v Basson 2005 (1) SALR 171 (CC)). 130

See Aust (n 28) 482–83 (discussing, among others, Garland v British Rail Engineering [1983] 2 AC 751). 131

See Sloss, ‗United States‘ (n 91) 526–27 (discussing, among others, Murray v Schooner Charming Betsy 6 US (2

Cranch) 64, 118 (1804)).

15

determine if there is any ambiguity, even latent, in the domestic legislation‘.132

Justice Kirby

advocated a similar approach in Australia, arguing that courts should refer to international

treaties ‗not only when there exists statutory ambiguity, but also where the construction of a

statute would result in an interpretation contrary to international human rights standards‘.133

However, the majority of the Australian High Court has rejected this approach, refusing ‗to

endorse a wider role for treaties in statutory interpretation other than where the legislature has

clearly envisaged such a role or where there exists a clear ambiguity on the face of the statute‘.134

Judicial application of the presumption is clearly transnationalist, especially in cases

where the statute is not facially ambiguous. In contrast, judges with a more nationalist orientation

sometimes avoid application of the presumption by declaring that a statute is unambiguous in

cases where litigants argue that the statute could reasonably be interpreted in conformity with

international treaty obligations.135

It is likely that courts throughout the world decide numerous

statutory interpretation cases where the presumption is not applied, even though it is potentially

applicable, because litigants fail to raise a possible treaty argument, or courts decline to address

the argument explicitly. It is difficult to perform a systematic analysis of judicial application of

the presumption even in a single country because it is hard to identify cases in which courts do

not mention potentially applicable treaty arguments.

B. Treaty Interpretation

Domestic courts in both monist and dualist states are frequently asked to interpret

treaties. In dualist states, this situation commonly arises when the legislature enacts a statute that

is expressly intended to implement a treaty.136

In monist states, courts sometimes interpret

treaties when a litigant asks the court to apply a treaty directly, and sometimes when the treaty is

applied indirectly. Regardless of the context in which treaty interpretation issues arise, courts

have a choice whether to adopt a nationalist or transnationalist approach to treaty interpretation.

Courts applying a transnationalist approach interpret treaties in accordance with the

shared understanding of the parties. In accordance with this approach, transnationalist judges cite

the Vienna Convention on the Law of Treaties,137

decisions of foreign courts138

and international

132

National Corn Growers Association v Canada [1990] 2 SCR 1324, 1372–73. 133

See Rothwell (n 24) 153–54. 134

ibid 156. 135

See, e.g., Breard v Greene 523 US 371, 375–76, 118 S Ct 1352, 1354–55 (1998) (construing a federal statute to

be inconsistent with U.S. obligations under the Vienna Convention on Consular Relations). 136

See, e.g., Aust (n 28) 482–83 (United Kingdom) (discussing, among others, Sidhu v British Airways [1997] 1

AER 193); Jayawickrama (n 16) 264–65 (India) (discussing Dadu alias Tulsidas v State of Maharashtra Supreme

Court of India, Writ Petition (Criminal) 169 of 1999, 12 October 2000); and van Ert (n 28) 177 (Canada) (discussing

Pushpanathan v Canada (Minister of Citizenship and Immigration) [1998] 1 SCR 982). 137

See, e.g., Aust (n 28) 483 (United Kingdom) (discussing, among others, R v Lambert Justices, ex p Yusufu [1985]

Times Law Reports 114); Garlicki and others (n 66) 387–89 (Poland) (discussing, among others, Decision of March

9, 2004, I CK 410/03 (not published Lex 182080)); Nollkaemper (n 87) 360–62 (Netherlands) (discussing, among

others, Supreme Court, State Secretary for Finance v X 21 February 2003, 36 NYIL 2005, 475); Rothwell (n 24)

151–52 (Australia) (discussing, among others, Morrison v Peacock [2002] HCA 44); and van Ert (n 28) 175–82

(Canada) (discussing, among others, Pushpanathan v Canada (Minister of Citizenship and Immigration) [1998] 1

SCR 982).

16

tribunals,139

and views adopted by non-judicial international bodies140

to support their

interpretations of particular treaty provisions. In contrast, courts applying a nationalist approach

emphasize that treaty interpretation is primarily an executive function, not a judicial function.

Accordingly, courts applying a nationalist approach tend to defer to the executive branch on

treaty interpretation issues.141

Deference to the executive branch often yields judicial opinions

that give greater weight to unilateral national policy interests, and less weight to the shared,

multilateral understanding that guides transnationalist interpretations.

Available information, which is admittedly limited, indicates that the nationalist approach

to treaty interpretation is a minority approach. The United States may be the only state where

courts have adopted an explicit interpretive presumption favoring deference to the executive

branch on treaty interpretation issues.142

In Israel, the Supreme Court has never adopted an

express interpretive presumption of this type, but ‗in cases relating to the [Occupied Territories],

for a long time, the Supreme Court in fact adopted the interpretation of [Geneva Convention IV]

favored by the authorities‘.143

In contrast, Polish commentators assert: ‗For a court to treat

executive branch views [on treaty interpretation issues] as dispositive would be incompatible

with the principle of independence of the judicial branch, as understood under the Polish

Constitution‘.144

The Polish view appears to be the dominant one. In most countries with

independent judiciaries — including both monist and dualist states — domestic courts claim an

138

See, e.g., Dugard (n 86) 470 (South Africa) (discussing, among others, Potgieter v British Airways 2005 (3)

SALR 133 (C)); Kretzmer (n 29) 291–92 (Israel) (discussing, among others, F.H. 36/84, Teichner v Air France 41

P.D. (1) 589); Nollkaemper (n 87) 364–65 (Netherlands) (discussing, among others, Administrative Law Division of

the Council of State, M.E.D. v State Secretary for Justice 6 November 1995, 28 NYIL 1997, 353); and van Ert (n 28)

185–86 (Canada) (discussing, among others, Connaught Laboratories Ltd v British Airways (2002) 61 OR (3d) 2004

(Ont. SCJ)). 139

See, e.g., Dugard (n 86) 466–70 (South Africa) (discussing, among others, Ferreria v Levin NO 1996 (1) SALR

984 (CC)); Garlicki and others (n 66) 389–98 (Poland) (discussing, among others, Judgment of January 29, 2003,

V.S.A. 1494/02 (ONSA 2004 nr 2, item 57)); Nollkaemper (n 87) 363–64 (Netherlands) (discussing, among others,

H v Public Prosecutor Court of Appeal of The Hague, ILDC 636 (NL 2007)); Paulus (n 79) 223–35 (Germany)

(discussing, among others, BVerfGE 111, 307 (2004)); and van Ert (n 28) 183–84 (Canada) (discussing, among

others, Mugesera v Canada (Minister of Citizenship and Immigration) [2005] 2 SCR 91). 140

See, e.g., Dugard (n 86) 466–70 (South Africa) (discussing, among others, Government of the Republic of South

Africa v Grootboom 2001 (I) SALR 46 (CC)); Minister of Health v Treatment Action Campaign (No 2) 2002 (5)

SALR 721 (CC)); Jayawickrama (n 16) 257 (India) (discussing Visaka v State of Rajasthan [1997] 3 LRC 361);

Kretzmer (n 29) 298–301 (Israel) (discussing, among others, HCJ 7029/95 New General Labor Federation v

National Labor Court 51 P.D. (2) 63, 157); and van Ert (n 28) 184–85 (Canada) (discussing, among others,

Canadian Foundation for Children, Youth and the Law v Canada (Attorney General) [2004] 1 SCR 76). 141

See, e.g., Sloss, ‗United States‘ (n 91) 524–25 (discussing, among others, El Al Israel Airlines, Ltd v Tsui Yuan

Tseng 525 US 155, 168 (1999)); Kretzmer (n 29) 292–95 (Israel) (discussing, among others, HCJ 785/87, Afu v IDF

Commander 42(2) P.D. 4 (English tr 29 Intl Legal Materials 139 (1990)). 142

See, e.g., Medellin v Texas 552 US 491, 513, 128 S Ct 1346, 1361 (2008) (‗It is, moreover, well settled that the

United States‘ interpretation of a treaty ―is entitled to great weight‖‘.). In other states, courts give conclusive weight

to executive views on specific treaty-related issues that are constitutionally committed to executive decision-making.

See, e.g., Aust (n 28) 484 (United Kingdom); Dugard (n 86) 471–72 (South Africa); and van Ert (n 28) 186–88

(Canada). However, to the best of my knowledge, the United States is the only country where the nation‘s highest

court has expressly adopted an interpretive canon favoring deference to the executive branch across the full

spectrum of treaty interpretation questions. 143

Kretzmer (n 29) 292 (discussing, among others, HCJ 785/87 Afu v IDF Commander 42(2) P.D. 4). 144

Garlicki and others (n 66) 399 (discussing, among others, K Galstyan V SA 726/99).

17

independent responsibility to construe treaties in accordance with the shared expectations of the

parties, without giving undue weight to the unilateral views of their own governments.145

A distinct interpretive issue concerns treaty-based protection for the rights of private

parties. Transnationalist judges recognize that many treaties are designed to protect the rights of

private parties. Accordingly, they interpret treaties in a manner that accords significant protection

to treaty-based private rights.146

In contrast, nationalist judges sometimes apply a presumption

that treaties ordinarily regulate horizontal relations between states, not vertical relations between

states and private parties.147

Application of this presumption can lead nationalist courts to

construe vertical treaty provisions as if they were horizontal provisions, thereby denying

protection for treaty-based private rights. This strategy provides nationalist judges a convenient

rationale for declining to apply treaty-based (vertical) constraints on governmental conduct.148

The United States is the only state whose courts have adopted an explicit interpretive

presumption that treaties do not create rights for private parties. Courts in other states approach

the matter as a straightforward interpretive question, without adopting a presumption for or

against private rights.149

If the treaty text, on its face, indicates that the parties intended to confer

rights on private parties, domestic courts will typically enforce those rights, subject to constraints

on judicial enforcement of unincorporated treaties in dualist states.150

C. Constitutional Interpretation

145

See, e.g., Aust (n 28) 482–83 (United Kingdom); Dugard (n 86) 471–72 (South Africa) (discussing, among

others, Kolbarschenko v King NO 2001 (4) SALR 336 (C)); Jayawickrama (n 16) 267–70 (India); Nollkaemper (n

87) 362–63 (Netherlands) (discussing, among others, Central Appeals Tribunal 21 July 2006, LJN No. AY 5560);

Paulus (n 79) 221–23 (Germany) (discussing, among others, Görgülü BVerfGE 111, 307, Engl Tr BVerfG, 2 BvR

1481/04 of 14 Oct 2004); and van Ert (n 28) 186–88 (Canada) (discussing Pushpanathan v Canada [1998] 1 SCR

982; Château-Gai Winers Ltd. v Attorney General of Canada [1970] Ex CR 366). 146

See, e.g., Aust (n 28) 484–87 (United Kingdom); Dugard (n 86) 472–74 (South Africa); Garlicki and others (n

66) 400–07 (Poland) (discussing, among others, Judgment of November 21, 2003 (I CK 323.02), OSNC 2004 nr 6,

item 103); Jayawickrama (n 16) 266–72 (India) (discussing Basu v State of West Bengal [1997] 2 LRC 1; Visaka v

State of Rajasthan [1997] 3 LRC 361); Nollkaemper (n 87) 345–48 (Netherlands) (discussing, among others,

Central Appeals Court for the Public Service and for Social Security Matters X Y & Z v B.O.Z. Regional

Compulsory Insurance Fund 29 May 1996, 30 NYIL 1998, 241); Paulus (n 79) 211 (Germany) (‗[W]hen individual

citizens claimed rights against the state on the basis of international law, it was quite natural that the state that had

given its word to other states could be regarded also bound toward its own citizens‘.); Rothwell (n 24) 136

(Australia) (‗[O]ne clear trend is that the courts have become more open to hearing matters based on the existence

not only of a treaty right recognized under Australian law but also of a right that exists entirely under international

law by way of a treaty to which Australia is a party‘.); and van Ert (n 28) 202–07 (Canada). 147

See Sloss, ‗United States‘ (n 91) 525–26 (discussing, among others, Gandara v Bennett 528 F.3d 823, 828 (11th

Cir. 2008)). 148

See ibid 539–40. 149

See, e.g., Aust (n 28) at 484 (United Kingdom) (‗There is no presumption that a treaty does not create a right for

a private party‘.); Dugard (n 86) 472 (South Africa) (stating that an incorporated treaty ‗creates rights and duties for

the individual in the same way that an ordinary statute creates rights and duties‘); Garlicki and others (n 66) 400

(Poland) (‗It is generally recognized that self-executing [treaty] provisions create rights (and obligations) for private

parties‘.); and Nollkaemper (n 87) 347 (Netherlands) (‗Dutch law recognizes the fact that states may agree by treaty

to grant certain rights to individuals, which they are then entitled to enforce before national courts‘.). 150

See supra notes 28-30 and accompanying text (discussing constraints on judicial application of unincorporated

treaties in dualist states.

18

Courts in both monist and dualist states apply treaties to help elucidate the meaning of

constitutional provisions. South Africa and India are two leading examples of states where courts

routinely invoke treaties and other provisions of international law in the context of constitutional

interpretation.151

The South African Constitution states explicitly: ‗When interpreting the Bill of

Rights, a court, tribunal or forum ... must consider international law; and may consider foreign

law‘.152

In light of this constitutional mandate, the South African Constitutional Court has

adopted the view ‗that the spirit, purport and objects of the bill of rights ... are inextricably linked

to international law and the values and approaches of the international community‘.153

Similarly,

the Indian Constitution stipulates: ‗The State shall endeavour to ... foster respect for international

law and treaty obligations in the dealings of organized peoples with one another‘.154

Accordingly, Indian jurisprudence reflects a view ‗that any international convention not

inconsistent with the fundamental rights provisions in the Constitution and in harmony with its

spirit must be read into those provisions to enlarge the meaning and content thereof‘.155

Courts in Canada, Germany, Israel and Poland also apply treaties to help interpret

domestic constitutional provisions, but they do so less regularly than the Indian Supreme Court

or the South African Constitutional Court.156

The judicial practice of using international law in

constitutional interpretation has provoked sharp controversy in both Australia and the United

States. In Australia, Justice Kirby was a strong advocate for judicial application of international

law in constitutional interpretation, but he never persuaded a majority of the High Court to

follow his recommended approach.157

The United States Supreme Court has occasionally cited

treaties to support its interpretation of a contested constitutional provision; in all such cases the

majority‘s reliance on international law provoked a strong critical response from the dissenting

Justices.158

Recent judicial practice in the United Kingdom merits separate discussion. Since Britain

does not have a written, constitutional Bill of Rights, British courts rely on other sources of law

to protect the fundamental rights that, in most other countries, are protected by a written

Constitution. The Human Rights Act, enacted in 1998, ‗effectively incorporated the [European

Convention of Human Rights] into English law‘.159

Since passage of the Act, British courts

151

See Dugard (n 86) 457–71 (South Africa) (discussing, among others, Prince v President Cape Law Society

2002(2) SALR 794 (CC)); Jayawickrama (n 16) 245–47, 266–72 (India) (discussing, among others, Visaka v State of

Rajasthan [1997] 3 LRC 361). 152

S. Afr. Const. s 39(1). 153

Dugard (n 86) 462 (quoting Neville Botha, ‗The Role of International Law in the Development of South African

Common Law‘ (2001) S. Af. Yrbk Intl L. 253, 259). 154

India Const. s 51. 155

Jayawickrama (n 16) 246 (discussing, among others Visaka v State of Rajasthan [1997] 3 LRC 361). 156

See Garlicki and others (n 66) 404–05 (Poland) (discussing, among others, Judgment of the Constitutional Court

of March 20, 2006 (K 17/05); Judgment of the Constitutional Court of October 11, 2006 (P 3/06)); Kretzmer (n 29)

298–301 (Israel) (discussing, among others, HCJ 112/77 Foguel v Broadcasting Authority 31 P.D. (3) 657); Paulus

(n 79) 230–33 (Germany) (discussing, among others, BVerfGE 111, 307 (2004)); and van Ert (n 28) 197–201

(Canada) (discussing, among others, R v Hape [2007] SCC 26). 157

See Rothwell (n 24) 156–58 (discussing, among others, Al-Kateb v Godwin [2004] HCA 37). 158

See, e.g., Graham v Florida 130 S Ct 2011 (2010); Roper v Simmons 543 US 551, 125 S Ct 1183 (2005); and

Lawrence v Texas 539 US 558, 123 S Ct 2472 (2003). For an insightful analysis of the controversy, see Mark

Tushnet, ‗International Law and Constitutional Interpretation in the Twenty-First Century: Change and Continuity‘

in Intl Law in the U.S. Supreme Court (n 114). 159

Aust (n 28) 487.

19

routinely apply the European Convention to protect individual rights that, in many other

countries, would be regarded as constitutional rights.160

D. Self-Execution in Monist States

Judicial doctrine in monist states distinguishes between treaties that are directly

applicable as law and treaties that are not directly applicable. Many states use the terms ‗self-

executing‘ and ‗non-self-executing‘ to distinguish between these two classes of treaty

provisions.161

When a court holds that a treaty is self-executing, it typically acts in a

transnationalist mode to facilitate the domestic application of treaty-based international norms.

When a court holds that a treaty is not self-executing, it generally acts in a nationalist mode to

shield the domestic legal system from the influence of treaty-based legal norms.162

Judicial

doctrine invariably grants judges some discretion to determine which treaties are self-executing.

Transnationalist judges exercise their discretion in a manner that pushes more treaties into the

self-executing category. Nationalist judges exercise their discretion in a manner that pushes more

treaties into the non-self-executing category.

South Africa‘s Constitution includes an explicit textual distinction between self-

executing and non-self-executing treaty provisions.163

Although the Constitution refers explicitly

to ‗self-executing‘ treaties, it does not define the term ‗self-executing‘, nor does it identify

criteria for distinguishing between self-executing and non-self-executing treaties. The South

African courts have not yet issued a definitive ruling to clarify the meaning of the self-execution

clause in the South African Constitution.164

Accordingly, there is an ongoing scholarly debate as

to which treaties, if any, are self-executing in South Africa.165

Ultimately, the resolution of that

question may have little practical significance because the South African Constitutional Court is

one of the most transnationalist courts in the world: it regularly applies treaties and customary

international law to help construe both statutory and constitutional provisions.166

Domestic courts in Germany, Poland, and the Netherlands are also fairly transnationalist,

insofar as they take a fairly broad view of which treaties are self-executing. In all three countries,

courts generally hold that treaty provisions designed to benefit private parties are invocable by

private parties and directly applicable by the courts, subject to one caveat.167

To be directly

160

See ibid 487–91 (discussing, among others, A(FC) and others v Secretary of State [2004] UKHL 56). 161

See, e.g., Dugard (n 86) 453–55 (South Africa); Garlicki and others (n 66) 400–03 (Poland); and Wildhaber and

others (n 69) 659 (Switzerland). 162

See Sloss, ‗United States‘ (n 91) 527–29 (discussing, among others, Intl Café SAL v Hard Rock Café Intl (USA),

Inc 252 F.3d 1274 (11th Cir. 2001)). 163

See S. Afr. Const. s 231(4), reprinted in Botha (n 68) 609–10 (‗Any international agreement becomes law in the

Republic when it is enacted into law by national legislation; but a self-executing provision of an agreement that has

been approved by Parliament is law in the Republic unless it is inconsistent with the Constitution or an Act of

Parliament‘.). 164

See Dugard (n 86) 454. 165

See ibid 453–55. 166

See ibid 457–73 (discussing, among others, S v Makwanyane 1995 (3) SA 391 (CC); S v Williams 1995 (3)

SALR 632 (CC)). 167

See Garlicki and others (n 66) 400–07 (Poland) (discussing, among others, The Interagra Judgment Judgment of

June 14, 1988 (II CR 367/87), OSP 1990 nr 9, item 705); Nollkaemper (n 87) 341–48 (the Netherlands) (discussing,

20

applicable, ‗a treaty provision has to be sufficiently clear to function as ‗objective law‘ in the

domestic legal order‘.168

Courts in all three countries have stated or assumed that most

substantive provisions of the European Convention on Human Rights and other human rights

treaties are self-executing.169

The self-execution jurisprudence in Germany, Poland and the Netherlands is

characteristic of most European Union countries because judicial decision-making in those

countries is heavily influenced by the European Court of Justice (ECJ). ECJ case law

―establishes that European law requires the direct effect of community law in the domestic legal

order. Moreover, the ECJ demands supremacy of European over domestic law.‖170

Thus, once a

legal instrument ―has been adopted by a competent EU body, it . . . becomes automatically

incorporated into the system of law binding on the national level [in Poland] and must be

enforced by all national authorities, in particular by the national courts.‖171

For states who are

members of the European Union, this is a ―consequence of EU membership,‖ and member states

have ―no alternative but to follow the established rules.‖172

In contrast to European jurisprudence, self-execution doctrine in the United States is

analytically incoherent.173

Courts and commentators agree that non-self-executing treaties are not

directly applicable by domestic courts, but they do not agree why this is so. Some sources

suggest that non-self-executing treaties are not incorporated into domestic law. A distinct view

holds that non-self-executing treaties are part of domestic law, but they are a special type of law

that courts are precluded from applying directly.174

Under the latter approach, there is further

disagreement as to why courts are precluded from applying non-self-executing treaties.175

In

among others, Central Appeals Tribunal, Management Board of Employee Insurance Benefits Agency v X 14 March

2003, 36 NYIL 2006, 466); and Paulus (n 79) 209–12 (Germany). 168

Nollkaemper (n 87) 333 (the Netherlands). See also Garlicki and others (n 66) 400 (courts in Poland recognize a

treaty as self-executing if it has been drafted ‗in a manner allowing its use as an exclusive legal basis for resolving

an individual case or controversy‘) (Resolution of the Supreme Court of October 5, 1974, III CZP 71/73, OSN CP

1975, no 5, item 72); Wildhaber and others (n 69) 659 (under Swiss doctrine, ‗a treaty is self-executing if it is

sufficiently precise and clear to constitute the basis of decision in a concrete case‘). 169

See Garlicki and others (n 66) 405 (Poland) (discussing Judgment of November 28, 2001 (V KKN 565/99, Lex nr

51620)); Nollkaemper (n 87) 342 (the Netherlands); and Paulus (n 79) 232 (Germany) (discussing BVerfGE 111,

307 (2004); English tr BVerfG, 2 BvR 1481/04 of 14 October 2004, available at

http://www.bverfg.de/entsheidungen/rs20041014_2bvri48104en.html). 170

Paulus (n 79) 210 (citing Internationale Handelsgesellschaft, Case 11/70, [1970] ECR 1125). 171

Garlicki and others (n 66) 385. 172

ibid 173

For recent commentary, see Curtis A. Bradley, ‗Self-Execution and Treaty Duality‘ (2009) Sup. Ct. Rev. 131;

Carlos Manuel Vazquez, ‗Treaties as Law of the Land: The Supremacy Clause and the Judicial Enforcement of

Treaties‘ (2008) 122 Harv. L. Rev. 599; and Ernest A. Young, ‗Treaties as ―Part of Our Law‖‘ (2009) 88 Tex. L.

Rev. 91. 174

See David Sloss, ‗Non-Self-Executing Treaties: Exposing a Constitutional Fallacy‘ (2002) 36 U.C. Davis L. Rev.

1, 10–18 (summarizing different theories of non-self-execution). 175

See ibid 12–18. In some cases, courts hold that a treaty provision is not self-executing because it is too vague or

ambiguous for judicial enforcement. See Carlos Manuel Vazquez, ‗The Four Doctrines of Self-Executing Treaties‘

(1995) 89 Am. J. Int‘l L. 695, 713-15. This version of the doctrine is similar to the non-self-execution doctrine

applied in many European countries. See supra notes 167-68 and accompanying text. In other cases, courts hold that

a treaty provision is not self-executing — even though it is sufficiently unambiguous to permit judicial enforcement

— because the political branches have manifested a desire to preclude or limit judicial enforcement. See Sloss (n

172) 35-44. This version of the doctrine has no apparent analogue in other countries.

21

practice, courts often hold that treaties are non-self-executing when an individual invokes a

vertical treaty provision as a constraint on government action, but they almost never hold that

transnational treaty provisions are non-self-executing.176

Thus, the net effect of judicial doctrine

is that U.S. courts tend to adopt a transnationalist approach in cases involving transnational treaty

provisions, but they tend to adopt a nationalist approach in cases involving vertical treaty

provisions.177

In contrast, courts in Germany, the Netherlands, Poland and South Africa adopt a

fairly consistent transnationalist approach for both vertical and transnational treaty provisions.

IV.

Domestic Courts and Treaty Compliance

The final part of this chapter addresses the respective roles of the judicial, executive and

legislative branches in promoting compliance with treaty obligations.178

My central claim is that

these roles vary greatly depending on whether the treaty provision at issue is horizontal, vertical

or transnational. In brief, executive officials have primary responsibility for ensuring compliance

with horizontal treaty obligations; the judiciary‘s role is marginal. With respect to transnational

treaty provisions, though, the positions are reversed. The judiciary plays a central role in

promoting compliance with transnational treaty provisions and the executive is marginalized.

The picture for vertical treaty provisions is more complex.

A. Horizontal Treaty Provisions

As discussed above, domestic courts rarely apply horizontal treaty provisions.179

Consequently, domestic courts bear little responsibility for promoting compliance with

horizontal treaty provisions. This proposition is generally true for both monist and dualist states,

regardless of whether courts adopt a nationalist or transnationalist approach.

With respect to horizontal treaties, the relationship between the legislative and executive

branches depends on the specific treaty provision at issue and the constitutional structure of a

given state. For example, the North Atlantic Treaty obligates parties to assist other member states

if there is ‗an armed attack against one or more of them in Europe or North America‘.180

The

duty to provide mutual assistance in the event of an armed attack is a paradigmatic horizontal

treaty obligation. If a NATO state was the target of an armed attack, the executive branches in

other NATO states would have primary responsibility for providing assistance under the treaty.

In some states, depending on constitutional separation of powers considerations, the executive

might have to obtain legislative approval before committing troops and weapons to the defense

of an ally. Regardless, there is no state in which the judiciary would be responsible for

implementing the nation‘s treaty obligation to help defend against an armed attack.

176

See Sloss, ‗United States‘ (n 91) 534–36. 177

See ibid 529–39. 178

Given space constraints, the present analysis does not address the role of sub-national governments. However, it

is noteworthy that sub-national governments in some states exert significant influence over matters related to treaty

compliance. See, e.g., Beemelmans & Treviranus (n 62) 328–29 (Germany); Wildhaber and others (n 69) 635–37

(Switzerland). 179

See supra notes 98-101 and accompanying text. 180

North Atlantic Treaty art 5 (adopted 4 April 1949, entered into force 24 August 1949) 34 UNTS 243.

22

B. Transnational Treaty Provisions

Conventional wisdom holds that the executive branch has primary responsibility in most

countries for implementing international treaty obligations. This is certainly not true for

transnational treaty provisions. Consider, for example, the 1929 Warsaw Convention, which

regulates international air carriage.181

In the United States, Congress never enacted legislation to

implement the Convention, but courts routinely apply it as a self-executing treaty.182

In many

dualist states,183

and even in some monist states,184

the legislature has enacted legislation to

promote effective implementation of the Convention. In all states — whether the treaty is

considered self-executing or is implemented by legislation — the judiciary bears primary

responsibility for resolving disputes between private parties that are governed by the

Convention.185

In the United States, the executive branch occasionally submits amicus briefs to

present its views about the proper interpretation of contested treaty provisions, but that is the

extent of executive branch participation in treaty implementation.

Domestic courts play a crucial role in promoting compliance with transnational treaty

provisions. A simple example helps illustrate this point. The New York Convention obligates

states to recognize and enforce foreign arbitral awards.186

Assume that a French company and a

Japanese company submit a commercial dispute to an arbitral panel in accordance with

UNCITRAL arbitration rules.187

The panel orders the Japanese company to pay damages to the

French company, but the Japanese company refuses to pay. That refusal, by itself, does not

constitute a violation of Japanese treaty obligations because the company‘s refusal to pay is not

attributable to the Japanese government.188

Now assume that the French company files suit in a

Japanese court to enforce the arbitral award. If the Japanese court rules against the French

company, and that ruling cannot be justified under the New York Convention,189

the judicial

decision would constitute a violation of Japanese treaty obligations because that judicial decision

is attributable to the Japanese government under principles of state responsibility.190

Conversely,

181

Convention for the Unification of Certain Rules Relating to International Carriage by Air (signed 12 October

1929, entered into force 13 February 1933) 137 LNTS 11 [hereinafter Warsaw Convention]. The Warsaw

Convention is the predecessor of the 1999 Montreal Convention (n 103). 182

See, e.g., Olympic Airways v Husain 540 US 644, 124 S Ct 1221 (2004); El Al Airlines v Tsui Yuan Tseng 525

US 155, 119 S Ct 662 (1999). 183

See, e.g., Kretzmer (n 29) 284 (Israel); Rothwell (n 24) 138 (Australia); and van Ert (n 28) 186 (Canada). 184

See Dugard (n 86) 470 (South Africa); Nollkaemper (n 87) 355 (the Netherlands). 185

See, e.g., Dugard (n 86) 470 (South Africa); Kretzmer (n 29) 290–91 (Israel); Nollkaemper (n 87) 355 (the

Netherlands); Rothwell (n 24) 138 (Australia); and van Ert (n 28) 186 (Canada). 186

See New York Convention (n 102) art. III (‗Each Contracting State shall recognize arbitral awards as binding and

enforce them ... under the conditions laid down in the following articles‘.). 187

UNCITRAL is the UN Commission on International Trade Law. For information, see UNCITRAL

<http://www.uncitral.org/uncitral/en/about_us.html> accessed 26 March 2011. 188

See James Crawford, The International Law Commission’s Articles on State Responsibility: Introduction, Text

and Commentaries (CUP, Cambridge 2002) 91–93. 189

Article V of the New York Convention identifies several circumstances in which ‗[r]ecognition and enforcement

of the [foreign arbitral] award may be refused‘. See New York Convention (n 102) art. V. Under the treaty, states

are obligated to enforce foreign arbitral awards unless there is a valid reasons for non-enforcement as specified in

Article V. 190

See Articles on Responsibility of States for Internationally Wrongful Acts art. 4 (‗The conduct of any State organ

shall be considered an act of that State under international law, whether the organ exercises legislative, executive,

judicial or any other functions ...‘) (reprinted in Crawford (n 186) 61).

23

if the Japanese court orders the Japanese company to pay — and especially if the court attaches

company assets to secure payment — the court is effectively acting as an agent of the

international legal system to ensure Japanese compliance with national treaty obligations. Either

way, the domestic court is the primary decision-maker whose decision determines whether the

nation complies with its treaty obligations. This is characteristic of transnational treaty

provisions: in most cases involving transnational provisions, domestic courts serve as the

primary interface between the domestic and international legal systems, and their decisions

effectively determine whether the nation complies with its treaty obligations.

The preceding observations about domestic judicial application of transnational treaty

provisions apply equally to both monist and dualist states, with one caveat. In dualist states, the

legislature typically incorporates a treaty before courts will apply it to resolve private disputes.

Once the treaty is incorporated, though, judicial application is quite similar in both monist and

dualist states. Moreover, the distinction between nationalist and transnationalist approaches has

scant effect on judicial application of transnational treaty provisions. The global record of

compliance with transnational treaty provisions is quite good because national courts in most

states apply transnational treaty provisions routinely — either directly or indirectly — to help

resolve private disputes arising from cross-border activities.

C. Vertical Treaty Provisions

The relationship among the legislative, executive and judicial branches in implementing

vertical treaty provisions is a complex subject that defies simple generalizations. Patterns vary by

nation and by individual treaty.

States sometimes achieve compliance with vertical treaty obligations even if no

government official or agency makes a conscious decision to implement that obligation. For

example, when the United States ratified the International Covenant on Civil and Political Rights

(ICCPR), the executive branch assured the Senate that no implementing legislation was

necessary because the United States could fulfill its treaty obligations by applying pre-existing

laws.191

Thus, when a court issues an injunction to enjoin enforcement of a state law that violates

federal laws prohibiting race-based discrimination, one could say that the court is promoting

compliance with U.S. treaty obligations under articles 2 and 26 of the ICCPR,192

even if the court

never considers a treaty-based argument. Similarly, commentators have noted that Canadian

courts implement Canada‘s obligations under the ICCPR, at least partially, by applying the

Canadian Charter of Rights and Freedoms and other provisions of domestic law.193

Leaving aside cases where states achieve compliance almost unwittingly, we turn next to

situations where some government actor makes a conscious decision to apply or interpret a

191

See David Sloss, ‗The Domestication of International Human Rights: Non-Self-Executing Declarations and

Human Rights Treaties‘ (1999) 24 Yale J. Intl L. 129, 183–88. 192

See ICCPR (n 3) art. 2(1) (guaranteeing protection of rights ‗without distinction of any kind, such as race, colour

... or other status‘); ibid art. 26 (guaranteeing ‗all persons equal and effective protection against discrimination on

any ground such as race, colour ... or other status‘). 193

See Gib van Ert, Using International Law in Canadian Courts (2nd

edn Kluwer Law International, The Hague

2008) 332–51.

24

particular treaty in a particular way. Here, it is helpful to discuss the 1951 Refugee Convention194

and the 1967 Refugee Protocol195

to illustrate the interplay among the legislative, executive and

judicial branches in the domestic application of vertical treaty provisions.

In dualist states, the legislature must first decide whether to enact legislation to

incorporate a treaty into domestic law. Professor Aust says: ‗It is invariable British practice

never to ratify a treaty until any [necessary implementing] legislation has first been made‘.196

Like Britain, other dualist states generally refrain from ratifying treaties with vertical obligations

unless or until they have enacted the implementing legislation necessary to ensure compliance

with those obligations.197

Accordingly, Australia, Canada, and the United Kingdom have all

adopted legislation to implement the Refugee Convention and Protocol.198

Even in monist states,

legislatures often enact implementing legislation to promote effective domestic implementation

of vertical treaty provisions. Although South Africa‘s Constitution provides expressly for self-

executing treaties,199

the South African legislature enacted legislation in 1998 to implement the

nation‘s treaty obligations under the Refugee Convention and Protocol.200

Similarly, in the

United States, even though the Constitution specifies that ratified treaties are the ‗supreme Law

of the Land‘,201

Congress enacted legislation in 1980 to implement U.S. obligations under the

Refugee Protocol.202

Thus, in both monist and dualist states, legislative decisions about whether

and how to implement vertical treaty provisions can have a significant impact on the nation‘s

compliance with its treaty obligations.

Once a vertical treaty provision enters into force domestically, the executive branch

assumes primary responsibility for treaty implementation. In most states, if an individual seeks

admission to the country as a refugee, an executive officer will make the initial determination

whether the individual qualifies for refugee status. That determination might promote or hinder

treaty compliance, depending on three factors: 1) whether the treaty has been fully or partially

incorporated into domestic law (either by legislation or self-execution); 2) insofar as the treaty is

unincorporated or partially incorporated, whether the executive decision-maker construes

relevant domestic laws in conformity with the nation‘s treaty obligations; and 3) insofar as the

decision-maker consults or applies the treaty, whether that decision-maker interprets the treaty in

accordance with internationally agreed principles of treaty interpretation.

194

Convention Relating to the Status of Refugees (adopted 28 July 1951, entered into force 22 April 1954) 189

UNTS 137. 195

Refugee Protocol (n 6). 196

Aust (n 28) 486. 197

See, e.g., Hollis ‗Comparative Approach‘ (n 20) 32–34 (concluding that ‗the case studies presented support a

practice by which states join treaties only after they have established the domestic legal means to comply with the

treaty‘s obligations‘); van Ert (n 28) 204 (‗The usual Canadian practice is not to allow treaties requiring domestic

implementation to enter into force for Canada until the federal government has ensured the treaty‘s

implementation‘.). 198

See Aust (n 28) 490–91 (United Kingdom); Rothwell (n 24) 138–40 (Australia); and van Ert (n 28) 175 (Canada). 199

See S. Afr. Const. (n 163). 200

See Dugard (n 86) 473–74. 201

U.S. Const. art. VI cl 2. 202

The Refugee Act of 1980, PL No. 96-212, 94 Stat. 102 (17 March 1980). The United States is not party to the

1951 Refugee Convention, but U.S. ratification of the 1967 Protocol effectively means that the United States is

bound by key provisions of the Convention.

25

If a treaty has been fully incorporated into domestic law — either by self-execution or by

legislative incorporation — the decision-maker will presumably apply the treaty as a rule of law

to reach his/her decision. In the Netherlands, for example, the 2000 Aliens Act authorizes

executive officers to grant residence permits for ‗Convention refugees‘, without defining the

term. Hence, the statute effectively directs administrative (and judicial) decision makers to apply

the treaty definition of refugees.203

The statute therefore promotes treaty compliance by directing

decision-makers to apply the treaty definition as a rule of domestic law. In contrast, when a

treaty remains wholly or partially unincorporated, decision-makers must apply domestic rules in

place of or in tandem with the international rule; this raises a greater risk of noncompliance. In

Australia, for example, the 1951 Convention has been only partially incorporated into domestic

law.204

Consequently, Australian decision-makers have been hesitant to rely too heavily on the

Convention in construing domestic statutes,205

producing a less-than-perfect record of treaty

compliance.

If a vertical treaty provision remains wholly or partially unincorporated, executive

decision-makers might still construe relevant domestic statutes in harmony with the nation‘s

international treaty obligations. For example, Canada‘s Immigration and Refugee Protection Act

directs executive officers to construe the Act ‗in a manner that ... complies with international

human rights instruments to which Canada is a signatory‘.206

The statutory reference to ―human

rights instruments‖ presumably includes the Refugee Convention and Protocol.207

Similarly, in

other states, executive officers may have a constitutional or statutory duty to perform their

governmental functions in a manner that is consistent with the nation‘s treaty obligations —

including, perhaps, obligations contained in unincorporated or partially incorporated treaties.208

Alternatively, executive officials might simply decide as a policy matter to exercise their

statutory responsibilities in a way that promotes compliance with treaty obligations. In any case,

if executive officials have a conscious goal of exercising their powers and duties consistently

with international treaty obligations, treaty compliance is enhanced. Conversely, if executive

officials are heedless of treaty obligations, their actions are less likely to promote treaty

compliance.

Executive officials are often required to interpret treaties. An official charged with

deciding whether to grant an applicant refugee status would need to interpret the treaty if the

treaty itself provides the governing rule of domestic law (via self-execution or full

incorporation), or if some law or policy directs the official to take account of the treaty when

construing the relevant domestic statute. In construing the treaty, the official might be guided to

203

See Nollkaemper (n 87) 336–37. 204

See Rothwell (n 24) 138–40. 205

See ibid 153–56. 206

van Ert (n 191) 155. 207

Canada is party to both the 1951 Refugee Convention and the 1967 Refugee Protocol. See United Nations Treaty

Collection <http://treaties.un.org/Home.aspx?lang=en> accessed 27 march 2011. 208

In the United States, the President and subordinate executive officers have a constitutional duty to ‗take Care that

the Laws be faithfully executed‘. U.S. Const. art. II s 3. Most commentators agree that the Take Care Clause creates

a legal duty for executive officers to implement treaties. See Edward T. Swaine, ‗Taking Care of Treaties‘ (2008)

108 Colum. L. Rev. 331. The U.S. Supreme Court decision in Medellin v. Texas, 552 US 491(2008), is not to the

contrary. Medellin stands for the proposition that there are constitutional limits on the President‘s power to

implement treaties, but Medellin is consistent with the view that the President has a duty to act within the scope of

his constitutional powers to execute treaties.

26

some extent by unilateral national policy interests. However, he or she might also be guided by

internationally agreed principles of treaty interpretation.209

If executive decision-makers give

great weight to internationally agreed principles, their decisions are more likely to promote treaty

compliance. Conversely, if decision-makers give more weight to unilateral policy interests, there

is a greater risk that their decisions will obstruct treaty compliance.

If the legislative and executive branches both viewed treaty compliance as a paramount

objective, the courts would rarely be asked to decide cases involving alleged treaty violations.

However, legislatures sometimes fail to implement treaties that require legislative

implementation, and executive officers sometimes fail to honor such treaties. When that happens,

courts may be asked to decide whether governmental conduct is consistent with the nation‘s

treaty obligations. Ultimately, the impact of judicial decision-making depends heavily on

whether domestic courts pursue a nationalist or transnationalist course. In states where courts

tend to adopt a transnationalist approach, domestic courts can play a key role in promoting treaty

compliance. India, the Netherlands, and Poland are leading examples of states where domestic

courts actively promote compliance with vertical treaty obligations.210

However, in states where

courts tend to apply a nationalist approach, domestic courts effectively cede authority to the

legislative and executive branches to make key decisions affecting compliance with vertical

treaty provisions. Israel and the United States exemplify this nationalist approach, although

judicial decision-making in Israel is moving in a more transnationalist direction.211

Finally, it is important to note that legislative action or inaction can nudge courts in a

more nationalist or transnationalist direction. In the United Kingdom, for example, Parliament‘s

decision to enact the Human Rights Act 1998 has undoubtedly moved judicial decision-making

in British courts in a more transnationalist direction.212

In the United States, however, the

Senate‘s consistent practice of attaching non-self-executing declarations to human rights treaties

has clearly pushed judicial decision-making in a more nationalist direction.213

These examples

illustrate the complexity of the relationship among legislative, executive and judicial branches in

shaping governmental decisions that affect compliance with vertical treaty obligations.

CONCLUSION

International law and international relations scholars have written extensively about

theories of national compliance with international legal obligations, including treaty

obligations.214

However, the scholarly literature has paid scant attention to domestic courts as

key institutional actors whose decisions can promote or impede treaty compliance.215

The

preceding discussion suggests that more detailed study of domestic courts is warranted. Granted,

209

Add citation to Richard Gardiner‘s chapter. 210

See Garlicki and others (n 66) (Poland); Jayawickrama (n 16) (India); and Nollkaemper (n 87) (the Netherlands). 211

See Kretzmer (n 29) (Israel); Sloss ‗United States‘ (n 91). 212

See Aust (n 28) 483–84, 487–90. 213

See, e.g., Martin S. Flaherty, ‗Global Power in an Age of Rights: Historical Commentary, 1946–2000‘ in Intl

Law in the U.S. Supreme Court (n 114). 214

See, e.g., Jack L. Goldsmith & Eric A. Posner, The Limits of International Law (OUP, Oxford 2005); Andrew T.

Guzmán, How International Law Works (OUP, Oxford 2008); and Ryan Goodman & Derek Jinks, ‗How to

Influence States: Socialization and International Human Rights Law‘ (2004) 54 Duke L.J. 621. 215

But see Harold Hongju Koh, ‗Why Do Nations Obey International Law?‘ (1997) 106 Yale L. J. 2599.

27

domestic judicial decisions have little impact on national compliance with horizontal treaty

obligations. However, domestic courts play a central role in ensuring compliance with

transnational treaty obligations. Moreover, domestic courts have the potential to play a very

significant role in promoting compliance with vertical treaty obligations. Whether that potential

is realized depends, to a great extent, on whether domestic courts adopt a nationalist or

transnationalist approach to the judicial application of vertical treaty provisions.